Several rotor systems were patented in the early 1920s and 30s. Edward Hebern, a native of Illionois, drew up a preliminary design for a rotor system in 1917. His company, Hebern Electric Code eventually failed. Alexander Koch filed his vision of a rotor cryptrographic machine in 1919 in post WWI Germany. His company would also fail. Arvid Gerhard Damm filed a patent in Holland a few days after Koch's for a similar rotor system. Of the three designs, Damm's was the least comprehensive patent filed. Damm's company would go on to make millions under new ownership by Boris Hagelin. All three patents, despite the differences in geography, are based on the same rotor concept and even use similar implementations.
The purpose of the rotor systems is to produce ciphertext using a large amount of cipheralphabets. Using many cipheralpahabets increases the period of a message that means that it will be unlikely that a cipheralphabet will be used twice in the same message. In paper and pen methods, the number of ciphertext alphabets is severely limited by the time it would take for a human to organize and use them effectively and accurately. Machines were the obvious next step in cryptography, and Hebern, Koch, and Damm all attacked the problem in similar ways.
The basic rotor system first created by all three employed a typewriter which was used for input, a series of rotor disks, and then either a printing mechanism or lightbulb system which displayed the ciphertext. When keys of the typewriter (representing plaintext), were pressed down, electricity flowed from the key, through the rotor disks, and out to a cipertext. The secret of creating the secure ciphertext came from the rotors.
Each rotor was a small disk, several inches in diameter and perhaps a half inch wide. The edges of the rotor were split into 26 equal sections and the letters of a randomly created ciphertext alphabet written in each space. Within the disk, a mess of wiring links each letter on the edge of the disk to another letter on another edge of the disk. The wiring could link one letter to the letter next to it, or to a letter on the other side of the disk. When a key was pressed, the electricity would flow from the plaintext to a letter on the edge of the rotor and then out to another letter on the rotor. Where the electricity emerged from the rotor would determine what the ciphertext letter was. With one stationary rotor, this system is nothing more than an elaborate monoalphabetic system.
If the rotor moves, though, the system suddenly becomes a polyalphabetic substitution. After a key is pressed the rotor can turn on space presenting different contacts. Pressing the same letter several times in a row would have resulted in the same letter with a stationary rotor, but if the rotor is moving after each letter is pressed, the ciphertext would be created from a different alphabet. On the 27th letter, however, the rotor would have made one complete revolution and the same cipheralphabet used for the first letter would be back in use. So, if you pressed the same letter 27 times, you would have the same letter on the 1st and 27th time. And if you pressed it again, you'd have a duplicate on the 2nd and 28th letters.
This means that the machine has a period of only 26 letters. A period that short is easily attainable using the original polyalphabetic systems by Porta and the price of such a mechanical one rotor moving device would be quite uneconomical. Improvements to this basic design made it the strongest pre-computer cryptography ever known.
